The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from a local file to inform its copywriting tasks, creating a potential vector for instructions hidden in marketing materials.
Ingestion points: The skill reads .claude/product-marketing-context.md as its primary source of project information.
Boundary markers: Absent; there are no specific delimiters or instructions for the agent to ignore directives that might be contained within the marketing context file.
Capability inventory: The skill possesses the ability to execute local shell commands via the bundled headline scorer script.
Sanitization: Absent; the content of the marketing context is processed and interpolated directly into the agent's workflow.
[COMMAND_EXECUTION]: The skill uses a bundled Python script to provide utility functions for the user.
Evidence: SKILL.md includes instructions for the agent to run python3 scripts/headline_scorer.py to evaluate marketing headlines.

copywriting